Our research continue to focus on objective response detection and threshold estimation using auditory evoked potentials (AEPs), with the recent addition of evoked otoacoustic emissions to our work. These physiological auditory responses are increasingly important, especially for detecting hearing loss in children, and become truly objective only when appropriate stimuli, response processing, decision variables, and statistical techniques are chosen. We propose to address several questions, all of immediate or potential clinical relevance: 1. How accurately and efficiently will our objective 40-Hz steady-state AEP (SSAEP) methods estimate behavioral thresholds in hearing-impaired adults? a. Will high-pass masking improve accuracy? 2. Are high-rate (greater than or equal to 80 Hz) SSAEPs really better than 40-Hz SSAEPs? a. How does state affect best rate? b. Do individuals have their own best rates? c. Will phase-weighting improve detection of high-rate SSAEPs, as for 40- Hz SSAEPs? 3. Will multi-frequency SSAEPs save time compared to sequential single- frequency SSAEPs? 4. Will SSAEPs elicited by narrow-band noise modulation of tones be more detectable than SSAEPs elicited by simple amplitude-modulated tones? 5. Will weighted averaging and phase-weighting improve detection of distortion-product otoacoustic emissions (DPOAEs)? a. Can the combination of phase-weighting and stimulus phase adjustment minimize the risk of passive artifactual (false-positive) DPOAE detection? 6. Will multiple-frequency simultaneous DPOAE measurement save time compared to sequential DPOAE measurement? 7. How well do selected time-domain (Fs, correlation) and frequency-domain (magnitude-squared coherence [MSC], phase coherence) statistics perform for detecting low level auditory brainstem responses? a. How large a latency estimation error will make phase-weighting ineffective in improving MSC performance? b. Should individual-frequency power estimates be averaged prior to, or after, calculation of MSC estimates? 8. Will the benefits of weighted averaging be greater when short (0.8 sec) subaverages are weighted, compared to longer subaverages (up to 13 sec)? a. Will spectral weighted averaging outperform conventional weighted averaging?